1. Field of the Invention
The present invention relates generally to an image generation method and device, and more particularly to a technology for combining one or a plurality of images sensed by an camera unit into an image whose sensing viewpoint is modified as if they were actually sensed from another viewpoint and displaying them.
2. Description of the Related Art
Generally, when monitoring using a monitor camera or the like, it is designed that an image sensed by a camera is displayed on a monitor, and an image sensed by a camera mounted in the desired place of a monitor area is displayed on a plurality of monitors arrayed in a monitor room. In order to contribute to safe driving, a camera is mounted on a vehicle, an area which an driver cannot directly or indirectly see is sensed using a camera directed toward the rear of the vehicle and such an image is displayed on a monitor provided in a driver seat.
However, since such a monitor device displays an image for a camera, a lot of cameras must be set in order to sense a wide area. If a wide-angle camera is used, the number of cameras decreases. However, since the accuracy of an image displayed on a monitor is rough, the displayed image is difficult to see and its monitor function degrades. In such a situation, a technology for combining a plurality of camera images into one image and displaying it is proposed. For example, in the following Reference 1, a plurality of camera images can be displayed in the divided areas of one monitor. In Reference 2, a plurality of cameras can be disposed in such a way that its sensed images can be partially overlapped, and it can be combined in its overlapped part into one image. In Reference 3, by combining the images of a plurality of cameras into one image by coordinate transformation, an image combined from an arbitrary viewpoint can be displayed.
In Reference 3, each piece of the image data of a plurality of cameras is one-dimensionally taken in and a three-dimensional space model is generated according to known information. Then, in this three-dimensional space model, this obtained image data is mapped according to the information of each pixel constituting an image inputted from a camera, based on its camera parameter to generate space data. After relating the images of all independent cameras to the three-dimensional space as a point in it in this way, a viewpoint-modified image which is viewed from an arbitrary virtual viewpoint instead of an actual camera viewpoint can be generated and displayed. According to such a viewpoint-modified image display method, the entire monitor area can be displayed from one arbitrary viewpoint without degrading image accuracy, and an area to monitor can be monitored from an arbitrary viewpoint.
Reference 1: Japanese Patent Application KOKAI Publication No. H05-310078
Reference 2: Japanese Patent Application KOKAI Publication No. H10-164566
Reference 3: Japanese Patent No. 3286306